Where do I begin?

Let me say that I need help analyzing a dual-balun problem with my short
loop. I have a configuration that is working great, but other than that
it shows that I know next-to-nothing about why it is working ...

I have a short wavelength (about 0.10 wavelength long) wire loop
attached to the balanced port eyelets of a 1:1 voltage balun from W2AU.

I am using twin-lead feedline but get this: the feedline is ALSO
attached to the the balanced port eyelets of the voltage balun. The
unbalanced port with the SO-239 jack is unused and unterminated. THIS
is driving me crazy since it seems to work.

At the end of my feedline I am connected to my tuner's 4:1 internal
voltage balun. (An old Swan/Cubic Tee-type tuner).

I definitely need the loop balun, otherwise my loop becomes a noisy
omnidirectional random wire. This wacky configuration seems to be the
only thing that keeps the whole system balanced, and I don't know why.

(I have also tried using a current balun (W2DU), but it also turns my
short directional loop into a random wire.)

It just plain looks wrong even though it works great. I'd appreciate
some feedback on my funky configuration. Since the antenna covers many
bands with the tuner, I make no attempt at matching the feedline
impedance to the loop, and use a tuner to resonate the whole system.

I've made a web-page about my projects in case you want to look at it
and laugh. My project site has info about my experiences with
coax-loops, but I have since moved on to plain wire loops since they
seem to perform as well or better than the coax-made ones.

htt://www.greertech.com/gpmag/gpmagloop.html

You can find the remote configuration at the bottom of the page.

I just can't explain what's going on here and could use your help. Many
thanks in advance!

73 de LoopFan

On Tue, 23 Mar 2004 02:28:32 GMT, loopfan wrote:

Where do I begin?

After reading this, that is what I asked myself too.

Let me say that I need help analyzing a dual-balun problem with my short
loop.

Small loop would be a better description. Also, loops are described
in diameters (or radius) not conductor length. Yes, there is a
correlation, and yes it is important to note the wavelength dimension
of that conductor length, but it is uncommon and most references stick
with diameter/radius for many and good reasons (such as radiation
resistance formulae).

I have a configuration that is working great, but other than that
it shows that I know next-to-nothing about why it is working ...

Which proves a little knowledge is dangerous I suppose, but we proceed
in good faith.

I have a short wavelength (about 0.10 wavelength long) wire loop
attached to the balanced port eyelets of a 1:1 voltage balun from W2AU.

Things were OK to this point.

I am using twin-lead feedline but get this: the feedline is ALSO
attached to the the balanced port eyelets of the voltage balun. The
unbalanced port with the SO-239 jack is unused and unterminated. THIS
is driving me crazy since it seems to work.

Yup, it caused me a whiplash too. I can't ask "why" you would want to
do this because you already admit you know next-to-nothing.

At the end of my feedline I am connected to my tuner's 4:1 internal
voltage balun. (An old Swan/Cubic Tee-type tuner).

That much makes sense - in isolation of the other statements.

I definitely need the loop balun

Considering you have described two BalUns, this reference is vague.

, otherwise my loop becomes a noisy
omnidirectional random wire.

You confirm one of the qualities of the BalUn. Obviously it
reinforces your hunch of doing something right (now if we can only
differentiate what that something is).

This wacky configuration seems to be the
only thing that keeps the whole system balanced, and I don't know why.

That is the same quality that rejected the noise: balance. This
balance through choking has eliminated common mode currents that blows
away the nulls by filling them in.

(I have also tried using a current balun (W2DU), but it also turns my
short directional loop into a random wire.)

How you did this is missing. If it was done in parallel with this
twin-lead, then you have what is called a ground loop, you shorted out
the choke. Again (with that "if") I won't ask why.

It just plain looks wrong even though it works great. I'd appreciate
some feedback on my funky configuration.

Fun-Kay!

Since the antenna covers many
bands with the tuner, I make no attempt at matching the feedline
impedance to the loop, and use a tuner to resonate the whole system.

All perfectly reasonable - in isolation of the other statements.

I've made a web-page about my projects in case you want to look at it
and laugh. My project site has info about my experiences with
coax-loops, but I have since moved on to plain wire loops since they
seem to perform as well or better than the coax-made ones.

htt://www.greertech.com/gpmag/gpmagloop.html

You (and everyone else) should cut and paste links from a browser
opened to the page that is referenced so we don't have the mistake
above. Typing URLs from memory often reveals IQ problems.

You can find the remote configuration at the bottom of the page.

I just can't explain what's going on here and could use your help. Many
thanks in advance!

73 de LoopFan

Hi OM,

The W2DU BalUn is an unnecessary item, dangling there like an
appendix. At most it is additional capacitance at the feed point.

Now, why you chose to switch from coax to twin-lead is a personal
issue. The distances involved do not demand it of necessity, that
much is certain. It cannot have anything to do with balance, because
your loop worked well with the coax and W2DU already. It works, which
is true enough, and enough to say it is simply a different way of
accomplishing the same thing.

As for the other material on your page about wire size, you need some
real data instead of those qualitative statements. I will repeat some
that I posted he
Let's look at small loops' Rr for various sizes in tabulated form:
Fo 1M diameter Efficiency with 1 mOhm loss
160M 29 µOhms 2.8%
80M 500 µOhms 33%
60M 1.5 mOhms 60%
40M 7.5 mOhms 88%
30M 24 mOhms 96%
20M 120 mOhms 99%

Let's examine the validity of that generous assignment of 1 mOhm loss
and see if it is reasonably warranted. Skin effect is the single
largest contributor to this loss as a source (aside from poor
construction techniques). Using the 1M diameter loop as being a
practically sized construction, and if we're using 2.54cM diameter
copper wire/tubing we find:
Fo skin effect loss
160M 13.8 mOhms
80M 20 mOhms
60M 23 mOhms
40M 28 mOhms
30M 33 mOhms
20M 39 mOhms

Well, 1 mOhm was too generous and if we look at those loops' Rr once
again against a robust, thick loop element:
Fo 1M diameter Efficiency with skin effect loss
160M 29 µOhms 0.2%
80M 500 µOhms 2.4%
60M 1.5 mOhms 6%
40M 7.5 mOhms 21%
30M 24 mOhms 42%
20M 120 mOhms 75%

Now, admittedly, this work above has a different diameter. but not
much different. It also reveals mediocre efficiency for 40M, but you
steer your readers away from that towards a much lossier system.

You might want to reflect on where the efficiency goes in your
suggested designs. Anecdotal reports such as yours suggest the #12
wire is fine, but the statement about large conductors being poor
sensitivity is suspect (you are suffering different issues).
Basically there is a conflict here you need to resolve.

By the way, flat strips are NOT better as substitutes for conductor
mass. If you must use aluminum-foil, wrap it around a hula hoop for
the round conductor shape that is much MORE efficient. You might want
to reflect how you are going to insure such aluminum-foil
constructions are going to maintain low mOhm conductivities - it
doesn't automatically come by wishing it so. Such foil oxidizes quite
rapidly on the unwaxed side (did you know there are two different
sides?). Needless to say, the waxed side isn't particularly
conductive either. Trying to solder to seams or make leads brings
even more issues spelled Ohm.

With small loops, the name of the game is resistance: pure and lossy.

73's
Richard Clark, KB7QHC

Richard,
It sounds like the balun at the loop's feed point is
adding inductance, not capacitance, wouldn't you think?
As for the 'why' of doing it that way, I can think of
several reasons, both intentional and not, ranging from
a "lets see what happens", to "oops, forgot to disconnect
the thing"...
'Doc

------------
| |
| |
| |
--oUUUUUUo--
| |
F F
L L

Not the best graphics in the world, but 'close'.

Hi Richard,

Wow - thanks for the feedback!

First off, I have totally removed the section about using twin-lead.
You are right, there is really no need for it considering the low
coaxial feedline cable loss at HF. I think I got tunnel-visioned on
some freak setup I found by accident.

How you did this is missing. If it was done in parallel with this
twin-lead, then you have what is called a ground loop, you shorted out
the choke. Again (with that "if") I won't ask why.

Yes, that is exactly what I did. Shorted choke - not good!


As for the other material on your page about wire size, you need some
real data instead of those qualitative statements. I will repeat some
that I posted he

Great stuff - I need to go back in and take away some of the more
misleading and/or subjective statements.

By the way, flat strips are NOT better as substitutes for conductor
mass. If you must use aluminum-foil, wrap it around a hula hoop for
the round conductor shape that is much MORE efficient. You might want

Ah, ok, so a round mass (pipe,tubing etc) is more efficient than a flat
strip even though they seem to share the same area? (ie a 6-inch wide
strip compared to a 3-inch diameter pipe) Is it kind of a shape vs. area
thing, assuming things like loss resistance are the same?

rapidly on the unwaxed side (did you know there are two different
sides?). Needless to say, the waxed side isn't particularly

I did not know that! I just thought the somewhat dull side was just due
to manufacturing and not due to wax. Geez, the more I learn the less I
know ....

Thanks again for your help - its most appreciated.

73 de WN6F
Brian

'Doc wrote:

As for the 'why' of doing it that way, I can think of
several reasons, both intentional and not, ranging from
a "lets see what happens", to "oops, forgot to disconnect
the thing"...

It was the "lets see what happens" situation. I brought the balun home
to test with a dipole, and got a flash of inspiration. Unfortunately, I
really needed eyelets on the input and the output of the balun and just
lashed it up.

I really didn't expect much but was amazed that it worked - and
frustrated at the same time because schematically it looked so bogus
having two baluns on each end of the feedline.

73 de WN6F

On Tue, 23 Mar 2004 09:58:40 GMT, loopfan wrote:
Ah, ok, so a round mass (pipe,tubing etc) is more efficient than a flat
strip even though they seem to share the same area? (ie a 6-inch wide
strip compared to a 3-inch diameter pipe) Is it kind of a shape vs. area
thing, assuming things like loss resistance are the same?

Hi Jack,

Basic physics of electrostatics teaches us that potential, in the form
of charge accumulates on a conductor at its smallest radius. Hence we
have lightning rods with sharp points, not flat metal plates occupying
the roof as a tile or shingle. The classic Van de Graff generator has
a dome construction whereby charge is moved up inside the dome, and it
immediately transfers itself to the outside surface (self shielding,
the interior has a negative radius in this sense). The dome serves
as a storage for the charge and presents the economy of a large radius
(the charge pump of the rubber belt presents a small radius inside
it). On the other hand, the Jacob's ladder consists of narrow wires
that emit continuous streams of arcing with much less voltage
(although at impressively high enough potential it is generally 1/10th
to 1/100th that of the Van de Graff).

A wire has an obvious radius (in cross section), and the charge is
distributed equally over its surface. However, if you hammer this
wire flat, the charge then seeks the edges (the smallest radius) and
abandons the flat area, starving it of conduction (resistance climbs).

The same phenomenon can be observed in variable capacitors that arc
further from their separated edges than from between their more
closely situated, meshed flat surfaces. Even with arcs between these
flat surfaces, it is always initiated by a site dislocality in the
form of a metal whisker or spur (small radius). Hence comes the
caution to finely polish the plates of high voltage capacitors.

It is a mistake think surface area alone as the geometry of a circular
cross section is more important.

73's
Richard Clark, KB7QHC

On Tue, 23 Mar 2004 03:34:47 -0600, 'Doc wrote:

Richard,
It sounds like the balun at the loop's feed point is
adding inductance, not capacitance, wouldn't you think?
As for the 'why' of doing it that way, I can think of
several reasons, both intentional and not, ranging from
a "lets see what happens", to "oops, forgot to disconnect
the thing"...
'Doc


Hi Doc,

A current balun design should be a simple short length of open coax
with ferrite beads = capacitor. This is one reason why it is benign
in this mis-application.

73's
Richard Clark, KB7QHC

Richard Clark wrote:

A wire has an obvious radius (in cross section), and the charge is
distributed equally over its surface. However, if you hammer this
wire flat, the charge then seeks the edges (the smallest radius) and
abandons the flat area, starving it of conduction (resistance climbs).

Wow, what a waste of material. Looks like flat strips will be put
waaaaay on the back-burner.

The same phenomenon can be observed in variable capacitors that arc
further from their separated edges than from between their more
closely situated, meshed flat surfaces. Even with arcs between these

Good example - all the caps I've seen that arced had them mostly near
the edges. I always wondered why.

It is a mistake think surface area alone as the geometry of a circular
cross section is more important.

I gotta' agree now. Thanks for the insight.

73 WN6F